Update scraper.py

scraper.py

@@ -2,54 +2,61 @@ import asyncio
 import random
 from curl_cffi.requests import AsyncSession
 
+# We use free CORS proxies to hide your Server IP from Cloudflare
+PROXIES = [
+    "",  # 1st Attempt: Direct Connection
+    "https://api.allorigins.win/raw?url=",  # 2nd Attempt: Proxy Router 1
+    "https://corsproxy.io/?",               # 3rd Attempt: Proxy Router 2
+]
+
 async def check_fragment(word: str) -> str:
-    url = f"https://fragment.com/username/{word}"
+    target_url = f"https://fragment.com/username/{word}"
     
-    # Impersonates a real Chrome browser to bypass Cloudflare
-    async with AsyncSession(impersonate="chrome110") as session:
-        for _ in range(4): 
+    headers = {
+        "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
+        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36"
+    }
+
+    # Spoof a real Google Chrome browser fingerprint
+    async with AsyncSession(impersonate="chrome120") as session:
+        for proxy in PROXIES:
+            url = f"{proxy}{target_url}"
+            
             try:
-                resp = await session.get(url, timeout=15)
+                resp = await session.get(url, headers=headers, timeout=15)
+                html = resp.text
                 
-                # Wait out rate limits
-                if resp.status_code in [429, 403]:
-                    await asyncio.sleep(random.uniform(2, 5))
-                    continue
-                    
+                # 1. Did we hit the Cloudflare Wall?
+                if resp.status_code in [403, 429] or "Just a moment..." in html or "cf-browser-verification" in html:
+                    proxy_name = "Proxy" if proxy else "Direct HF IP"
+                    print(f"⚠️ [CLOUDFLARE BLOCK] Blocked on {proxy_name} for @{word}. Retrying...")
+                    await asyncio.sleep(random.uniform(2, 4))
+                    continue # Try the next proxy in the list
+                    
+                # 2. Success! Parse the HTML accurately
                 if resp.status_code == 200:
-                    html = resp.text
                     
-                    # Double-check if Cloudflare intercepted the page anyway
-                    if "Just a moment..." in html or "Cloudflare" in html:
-                        await asyncio.sleep(random.uniform(3, 6))
-                        continue
-                        
-                    await asyncio.sleep(random.uniform(0.15, 0.6))
-                    
-                    # --- HTML PARSER ---
-                    if "tm-status-avail" in html: return "AVAILABLE"
+                    if "tm-status-avail" in html: 
+                        print(f"✅ [SUCCESS] @{word} is AVAILABLE!")
+                        return "AVAILABLE"
                     
                     if "tm-status-taken" in html:
                         if "On auction" in html or "tm-section-subscribe" in html: return "ON_AUCTION"
                         if "For sale" in html: return "FOR_SALE"
                         if ">Sold for" in html or ">Sold at" in html or 'class="tm-status-text">Sold' in html: return "SOLD"
                         return "TAKEN"
-                    
-                    if "tm-status-unavail" in html: return "UNAVAILABLE"
+                        
+                    if "tm-status-unavail" in html: 
+                        return "UNAVAILABLE"
 
-                    # Fallbacks
-                    if "Available" in html: return "AVAILABLE"
-                    if "Taken" in html: return "TAKEN"
-                    
-                    # If Fragment redirects to the search page, the word is usually available to claim
-                    if "/?query=" in str(resp.url):
-                        return "AVAILABLE"
-                    
-                    return "UNAVAILABLE"
+                    # If Fragment loads but the layout changed
+                    print(f"❓ [UNKNOWN HTML] Couldn't parse @{word}. Fragment may have changed their website.")
+                    return "ERROR"
                     
-            except Exception:
-                pass 
-            
-            await asyncio.sleep(random.uniform(1.0, 2.0))
-            
-        return "UNAVAILABLE"
+            except Exception as e:
+                print(f"❌ [NETWORK ERROR] Failed to load @{word}: {str(e)}")
+                continue 
+                
+    # If all proxies fail, DO NOT label as Unavail. Label as ERROR to protect the database.
+    print(f"💀 [FATAL] All proxies blocked for @{word}.")
+    return "ERROR"